flapi

Name: flapi
Rating: 2.3 (70 reviews)
Author: DataZooDE

by DataZooDE·★ 70·Score 46

flAPI generates SQL-based APIs with DuckDB and provides MCP tools for AI agents.

databasedeveloper-toolsai-llm

Forks

Open issues

this month

Last commit

2d ago

Indexed

Overview

flAPI is a C++ framework that creates read-only APIs from SQL templates using DuckDB. It offers unified configuration for both REST endpoints and MCP tools, allowing developers to expose database connections through APIs with automatic validation and caching. The framework supports multiple data sources including BigQuery, Snowflake, Parquet, Iceberg, and various SQL databases, with built-in security features and Swagger documentation.

Try asking AI

After installing, here are 5 things you can ask your AI assistant:

you:Creating APIs for large datasets without writing custom backend code

you:Building data tools for AI agents with SQL-based access

you:Exposing enterprise data sources as secure endpoints with authentication

you:What data sources does flAPI support?

you:How is MCP configured in flAPI?

When to choose this

Choose flAPI when you need to quickly create APIs for existing data warehouses or databases that AI agents need to access, especially when you're already using DuckDB or want read-only access without ETL.

When NOT to choose this

Don't choose flAPI if you need write access to your databases, require complex business logic beyond SQL queries, or need to connect to data sources not supported by DuckDB.

Tools this server exposes

1 tool extracted from the README

get_customers
Retrieve customer information by ID

Note: Tool name inferred from unified configuration example where mcp-tool: get_customers is defined

Comparable tools

duckdb-mcppostgres-mcpsupabase-mcpbigquery-shell-mcp

Installation

The easiest way to try flAPI is using uvx:

uvx --from flapi-io flapi -c flapi.yaml

Or install permanently:

pip install flapi-io

Using Docker:

docker pull ghcr.io/datazoode/flapi:latest
docker run -it --rm -p 8080:8080 -p 8081:8081 -v $(pwd)/examples/:/config ghcr.io/datazoode/flapi -c /config/flapi.yaml --enable-mcp

Claude Desktop Configuration

Add to your Claude Desktop config.json:

{
  "mcpServers": {
    "flapi": {
      "command": "flapi",
      "args": ["--enable-mcp"]
    }
  }
}

FAQ

What data sources does flAPI support?: flAPI supports BigQuery, Snowflake, Parquet, Iceberg, Postgres, MySQL, SAP ERP & BW, and other data sources through DuckDB extensions.
How is MCP configured in flAPI?: MCP is enabled either with the --enable-mcp flag or by setting 'mcp: enabled: true' in the flapi.yaml configuration file.

Compare flapi with

flapi vs ultimate_mcp_server flapi vs mcp-server-chart flapi vs everything flapi vs filesystem flapi vs time

GitHub →

Last updated 2026-05-17 · Auto-generated from public README + GitHub signals.